Effects of binary mixtures of xenoestrogens on gonadal development and reproduction in zebrafish

Aquatic Toxicology 80 (2006) 382–395

Effects of binary mixtures of xenoestrogens on gonadal development and reproduction in zebrafish Leo L. Lin a , David M. Janz a,b,∗ a

b

Toxicology Centre, University of Saskatchewan, Saskatoon, SK, Canada S7N 5B3 Department of Veterinary Biomedical Sciences, Western College of Veterinary Medicine, University of Saskatchewan, Saskatoon, SK, Canada S7N 5B4 Received 15 September 2006; received in revised form 10 October 2006; accepted 11 October 2006

Abstract Previous studies exposing fish to xenoestrogens have demonstrated vitellogenin (VTG) induction, delayed gametogenesis, altered sex ratio, and decreased reproductive performance, with a majority of those studies focusing on exposure to single chemicals. The objective of this study was to determine the effects of binary mixtures of a weak estrogen receptor agonist, nonylphenol (NP) and a potent estrogen receptor agonist, 17␣-ethinylestradiol (EE) on sex ratios, gametogenesis, VTG induction, heat shock protein 70 (HSP70) expression and reproductive capacity in zebrafish (Danio rerio). Fish were exposed from 2 to 60 days post-hatch (dph) to nominal concentrations of 10 or 100 ␮g/l NP (NP10 or NP100, respectively), 1 or 10 ng/l EE (EE1 or EE10, respectively), 1 ng/l EE + 10 or 100 ␮g/l NP (EE1 + NP10 or EE1 + NP100, respectively), 10 ng/l EE + 10 or 100 ␮g/l NP (EE10 + NP10 or EE10 + NP100, respectively) or solvent control (0.01% acetone, v/v) in a static-renewal system with replacement every 48 h. At 60 dph, fish from each treatment were euthanized for histological examination of gonads, and whole body VTG and HSP70 levels. Remaining fish were reared in clean water until adulthood (240 dph) for breeding studies. In all EE10 exposure groups (EE10, EE10 + NP10 and EE10 + NP100), increasing NP concentration acted antagonistically to the action of EE in terms of VTG induction at 60 dph. Similarly, non-additivity was observed with egg production, where EE1 + NP100 exposure resulted in significantly more eggs produced per breeding trial than EE1 alone. Histological staging of oogenesis revealed suppressed gametogenesis in an additive fashion in females at 60 dph. There were no differences among treatment groups in whole body HSP70 expression in 60 dph fish or in gonadal HSP70 expression in adult fish. Although there was no statistical evidence of non-additivity, breeding trials in adults revealed significant reductions in egg viability, egg hatchability and/or F1 swim-up success, suggesting that developmental exposures to xenoestrogens may cause irreversible effects on egg quality and progeny even after periods of depuration. In conclusion, these results suggest that environmentally relevant mixtures of NP and EE can produce additive or non-additive effects that depend on the particular response being determined and the respective exposure concentrations of each chemical. © 2006 Elsevier B.V. All rights reserved. Keywords: Ethinylestradiol; Nonylphenol; Zebrafish; Vitellogenin; Reproduction; Mixture

1. Introduction The last decade has witnessed a dramatic increase in research efforts devoted to the occurrences and effects of xenoestrogens in the environment. Xenoestrogens are compounds that exhibit the ability to mimic, antagonize or alter action of the endogenous estrogen, 17␤-estradiol (E2 ), primarily by interacting with estrogen receptors (ERs) (Danzo, 1997). Numerous studies have reported that exposure to such compounds may disturb normal ∗

Corresponding author at: Department of Veterinary Biomedical Sciences, Western College of Veterinary Medicine, University of Saskatchewan, 52 Campus Drive, Saskatoon, SK, Canada S7N 5B4. Tel.: +1 306 966 7434; fax: +1 306 966 7376. E-mail address: [email protected] (D.M. Janz). 0166-445X/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.aquatox.2006.10.004

endocrine function in animals and humans, leading to various developmental and reproductive impairments (Colborn et al., 1993; Sumpter, 1998; Guillette and Gunderson, 2001). In fish, where a majority of studies have focused, xenoestrogen exposures led to changes in plasma hormone concentrations (Khan and Thomas, 1998), gonadal size (Jobling et al., 1996; Gray and Metcalfe, 1997; Ashfield et al., 1998), development of intersex gonads (ovotestis) (Gray and Metcalfe, 1997), and induction of the female yolk precursor protein, vitellogenin (VTG) (Jobling et al., 1996). Anthropogenic compounds that have been identified as xenoestrogens encompass a wide range of chemical structural diversity and include organochlorine pesticides, polychlorinated biphenyls, dioxins, alkylphenols, and phthalates (Allen et al., 1999). In comparison to E2 , the estrogenic potency of most of

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

these chemicals is low since they bind to ERs with much less affinity, however, they can occur at concentrations in the environment that are sufficient to induce some typical biological effects of E2 (Jobling and Sumpter, 1993; Guillette et al., 1994; Soto et al., 1995; Routledge and Sumpter, 1996). Amongst xenoestrogens, alkylphenols are one of the more prevalent and environmentally persistent. Alkylphenol polyethoxylates (APEOs) are North America’s second largest group of nonionic surfactants in commercial production. They are widely used in a variety of industrial, agricultural, and household applications. Of the estimated 0.3 Mt year−1 APEOs produced worldwide, about 80% are nonylphenol polyethoxylates, while octylphenol polyethoxylates account for most of the remaining 20% (Naylor et al., 1992). During sewage treatment, biodegradation of nonylphenol polyethoxylates occurs by the hydrolytic removal of ethoxylate groups into short-chain ethoxylates, carboxylic acid derivatives, and eventually to nonylphenol (NP) (Nimrod and Benson, 1996). Concentrations of NP in final effluents have been reported to range from non-detectable to 330 ␮g/l (Bennie, 1999). In addition to the aforementioned synthetic chemicals, recent studies have demonstrated, in sewage treatment effluents and surface waters, presence of the endogenous estrogen E2 and its metabolites estrone and estriol, as well as the synthetic 17␣ethinylestradiol (EE). Ethinylestradiol is one of the most commonly used active ingredients in oral contraceptives (ArcandHoy et al., 1998). It is normally excreted in the urine as inactive glucuronide conjugates, but is readily activated through microbial ␤-glucuronidase activity during wastewater treatment (Guengerich, 1990; Desbrow et al., 1998). In Europe and North America, EE has been detected in many sewage treatment plant effluents and surface waters, with concentrations ranging from nondetectable to 62 ng/l in effluents (Desbrow et al., 1998; Ternes et al., 1999; Williams et al., 2003), and from below detection limit to 15 ng/l in surface waters (Aherne and Briggs, 1989; Jobling et al., 1996; Belfroid et al., 1999; Ternes et al., 1999) being reported. Determination of VTG is one of the most frequently used in vivo exposure biomarkers for estrogenicity. Vitellogenesis is an ER-mediated response whereby VTG is produced in the liver and transported via blood to the ovaries where it is incorporated into the developing oocytes. Normally VTG concentration is very low or absent in male and juvenile fish, but can be induced upon xenoestrogen exposure (Sumpter and Jobling, 1995; Folmar et al., 1996; Jobling et al., 1996; Tyler et al., 1998). Induction of heat shock proteins (HSPs), also known as stress proteins, have also been proposed as biomarkers of toxicant exposure in aquatic species (Iwama et al., 1998). The 70 kDa stress protein, HSP70, is probably the best characterized and studied. Varieties of metals (Ryan and Hightower, 1994; Williams et al., 1996; Yoo and Janz, 2003), bleached kraft pulp mill effluents (Janz et al., 1997; Vijayan et al., 1998; Janz et al., 2001), and aryl hydrocarbon receptor (AhR) agonists (Weber and Janz, 2001) have been reported to induce HSP70 expression in fish. Zebrafish (Danio rerio) was chosen as the test species for this study because of its small size, ease of maintenance and

383

short reproductive cycle. It is an undifferentiated gonochorist, where genotypic males go through a period of transitory hermaphroditism during the juvenile developmental stage (Takahashi, 1977). All gonads of zebrafish begin with development of ovary-like tissues. Soon afterwards, future females continue to develop ovaries, while in future males, the ovarian tissues degenerate and disappear, followed by the development of testicular tissues. Full sex reversal and testicular formation occurs by approximately 40 dph (Takahashi, 1977; Uchida et al., 2002). Several studies have reported that exposing zebrafish to xenoestrogens during the period of sexual differentiation results in altered sex ratios, suppressed gonad development, induction of VTG, and development of ovotestes in zebrafish (Andersen et al., 2003; Hill and Janz, 2003; Orn et al., 2003). Thus far, the majority of research on effects of xenoestrogens on aquatic species has focused primarily on exposure to single chemicals, despite the fact that they are more likely to be exposed to combinations of such compounds in the environment. Although there has been an increased effort to understand mixture toxicity of xenoestrogens (Thorpe et al., 2001, 2003; Brian et al., 2005), few studies have investigated in vivo effects of xenoestrogen mixtures on early development and subsequent reproduction. The objectives of the present study were to determine the effects of developmental exposure to binary mixtures of EE and NP on VTG induction, HSP70 expression, gametogenesis, sex ratios, and reproductive fitness of zebrafish. Our null hypothesis was that the effects of binary mixtures of EE and NP on these responses would be simply additive. 2. Materials and methods 2.1. Test compounds 17␣-Ethinylestradiol (17␣-ethinyl-1,3,5[10]-estratriene3,17␤-diol; 98% purity) and 4-nonylphenol (technical grade) were obtained from Sigma–Aldrich (St. Louis, MO, USA). Stock solutions of EE and NP were prepared in HPLC-grade acetone. Solvent concentration was kept at 0.01% (v/v) throughout the experiment. 2.2. Experimental animals Adult zebrafish for breeding stock were purchased from a local supplier and housed in a temperature (28 ± 1 ◦ C) and photoperiod (16 h light:8 h dark) controlled environmental chamber. Approximately 200 fish were divided evenly between ten 40-l glass aquaria supplied with dechlorinated tap water (pH: 7.7; conductivity: 380 ␮S/cm; hardness: 128 mg/l CaCO3 ; alkalinity: 78 mg/l CaCO3 ; dissolved organic carbon: 2.5 mg/l; total dissolved solid: 210.8 mg/l). Aeration and filtration were provided by Biofoam sponge filters (Hagen, Montreal, Que., Canada). Fish were fed twice daily with Nutrafin Max flake food (Hagen, Montreal, Que., Canada) in the morning, supplemented with freshly hatched brine shrimp (Artemia nauplii) in the evening. Fish were acclimated to laboratory conditions for 4 weeks prior to breeding.

384

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

2.3. Parental breeding

2.6. Gametogenesis

Adults were bred to obtain eggs for the exposure experiments. Plastic spawning trays covered by meshed lid were placed in tanks in the afternoon of the day prior to breeding. On the following day, at 1 h after the start of the light cycle, spawning trays were removed from the tanks. Eggs were collected and placed into sterile salt-based egg water and assessed for viability under a dissecting microscope. The egg water was prepared by dissolving a 60 g/l solution of instant ocean sea salts in 1 l ddH2 O (Westerfield, 1995). Viable eggs were rinsed thoroughly with egg water and held in sterile Petri dishes until hatch.

Staging of oogenesis and spermatogenesis was conducted following the quantitative method described by Weber et al. (2003). Staging of ovarian development was performed on sections of histologically verified female fish. Using an Olympus AH-2 light microscope at 200× magnification, the number of ovarian follicles at oogonial (smallest in size, with relatively larger nucleus contained in eosinophilic ooplasm) and previtellogenic (small in size, basophilic ooplasm and large nucleus with visible chromatin and single somatic cell layer) stages were enumerated. The percent of follicles at each developmental stage was calculated as a percent of the total number of follicles in each view. Four replicate views were performed for each fish. Staging of testicular development was performed on sections of histologically verified male fish. Using an Olympus AH-2 light microscope at 1000× magnification, the number of spermatocysts containing spermatogonia (eosinophilic cytoplasm with relatively large nucleus), primary or secondary spermatocytes (thread-like or condensed chromatin, respectively, with relatively smaller cytoplasm that does not take up dye) and spermatids or mature sperm (tightly packed nuclear material lacking surrounding cytoplasm and with a developed tail) stages were enumerated. The percent of spermatocysts at a given developmental stage was calculated as a percent of the total number of spermatocysts in each view. Four replicate views were performed for each fish.

2.4. Exposure assays Newly hatched fry were held in sterile Petri dishes (50 fry per dish) containing egg water. At 2 days post-hatch (dph), fry were exposed to nominal concentrations of EE (1 or 10 ng/l; denoted EE1 or EE10), NP (10 or 100 ␮g/l; denoted NP10 or NP100), EE + NP (1 ng/l + 10 ␮g/l, 1 ng/l + 100 ␮g/l, 10 ng/l + 10 ␮g/l, or 10 ng/l + 100 ␮g/l; denoted EE1 + NP10, EE1 + NP100, EE10 + NP10, or EE10 + NP100) or acetone solvent (control) at a 1 ␮l/ml total dilution in system water. There were three replicate aquaria for each test chemical concentration and solvent control. A 100% water change was performed every 48 h from 2 to 60 dph. Fresh acetone or test chemicals were added at the time of each water change. From 2 to 30 dph, fry were fed an alternating diet of Paramecium multimicronucleatum or freshly hatched brine shrimp three times daily. At 7 dph, fry were transferred to aerated 250 ml beakers. At 30 dph, fry were transferred to aerated 1 l glass beakers, and an alternating diet of flake food and newly hatched brine shrimp was provided twice daily. Chemical exposures continued until 60 dph. At 60 dph, 42 fish (14 fish per replicate) from each treatment group were randomly selected for measurement of lengths, weights, histological examination of the gonads, HSP70 and VTG determinations. Remaining fish were transferred to 20 l glass aquaria and reared in dechlorinated municipal tap water to allow a 6 month depuration period until 240 dph, at which time breeding trials were conducted. 2.5. Histology At 60 dph, 21 (7 from each of the 3 replicates) fish from each treatment were euthanized with an overdose of MS-222 (3-aminobenzoic acid ethyl ester, methanesulfonate salt). Body weights and lengths were recorded, and condition factors were calculated by [body weight (g)/length (mm)3 ] × 100,000. Fish were then fixed in Bouin’s solution for 24 h and subsequently transferred to 70% ethanol. Whole fish were dehydrated through a graded series of ethanol and embedded in paraffin wax. Longitudinal 5 ␮m sections along the entire dorso-ventral axis were taken with a microtome at 20 ␮m increments, collected onto glass slides, and stained using hematoxylin and eosin. Slides were then analyzed blind of treatment with a light microscope to evaluate the presence or absence of gonad tissue and determine the phenotypic sex.

2.7. Sixty days post-hatch whole body vitellogenin determination The remaining 60 dph fish were stored at −80 ◦ C until determinations of VTG and HSP70. Vitellogenin levels were determined using a commercial ELISA kit for zebrafish VTG (Biosense Laboratories, SA, Bergen, Norway) in 60 dph zebrafish whole body homogenates, following the instructions of the manufacturer. Homogenate samples were directly applied to the assay at appropriate dilutions. Vitellogenin concentrations were normalized to the weight (g) of the corresponding sample, and are expressed as mg VTG per g body weight. 2.8. Sixty days post-hatch whole body HSP70 determination Whole body homogenates of 60 dph zebrafish were prepared to determine HSP70 expression using Western blotting. Individual fish were homogenized in 200 ␮l ice cold buffer (50 mM HEPES, 150 mM NaCl, 1 mM EGTA, 1.5 mM MgCl2 , 1% (v/v) Triton X-100, 10% (v/v) glycerol, pH 7.5) containing protease inhibitors (0.1 mg/ml 2-aminoethyl-benzenesulfonyl fluoride (AEBSF), 20 ␮g/ml soybean trypsin inhibitor, and 1.9 ␮g/ml aprotinin). Samples were gently mixed for 1 h at 4 ◦ C, and then centrifuged for 25 min at 5200 rpm and 4 ◦ C. Following centrifugation, the supernatant was withdrawn and stored at −80 ◦ C. The protein concentration of each sample was determined (DC Protein Assay, BioRad, Hercules, CA, USA) using bovine serum albumin as standard. Proteins (30 ␮g protein per lane) were

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

385

Table 1 3 × 3 two-way ANOVA matrix for evaluating the interaction effects between EE and NP on the examined endpoints

NP0 NP10 NP100

EE0

EE1

EE10

EE0NP0 (control) EE0NP10 (NP10) EE0NP100 (NP100)

EE1NP0 (EE1) EE1NP10 (EE1 + NP10) EE1NP100 (EE1 + NP100)

EE10NP0 (EE10) EE10NP10 (EE10 + NP10) EE10NP100 (EE10 + NP100)

separated using 10% SDS-PAGE and transferred to 0.45 ␮m nitrocellulose membranes (BioRad) for 17 h at 4 ◦ C. HSP70 was detected with a monoclonal mouse anti-bovine HSP70 antibody (Sigma–Aldrich H5147) at 1:5000 dilution. Blots were visualized with 1:2000 horseradish peroxidase-conjugated goat antimouse antibody (Santa Cruz Biotech, Santa Cruz, CA, USA) and NBT/BCIP color development. HSP70 immunoreactive bands were quantitated by densitometry using an Epson 4180 scanner and Scion image beta (Version 4.02) software.

test was used followed by multiple comparisons versus the respective control groups (Dunn’s method). Chi-square analysis was used to identify differences in sex ratio between control and each treatment group. Statistical significance was set at α = 0.05.

2.9. Reproductive studies at 240 dph

Mean survival rates from 2 to 60 dph were 77.3 ± 6.4% (CON); 77.3 ± 7.0% (NP10); 61.3 ± 22.3% (NP100); 83.3 ± 9.9% (EE1); 78.7 ± 9.9 (EE10); 61.3 ± 18.6 (EE1 + NP10); 74.0 ± 4.0% (EE1 + NP100); 74.7 ± 16.2% (EE10 + NP10) and 88.0 ± 5.3% (EE10 + NP100). There were no significant differences in the 60 dph survival rate among treatments (data not shown).

Upon completion of the exposure study at 60 dph, the remaining fish were reared in dechlorinated municipal tap water and allowed a 6 month depuration period, at which time the reproductive study commenced. Three replicates of 15 breeding trials were conducted. For each replicate, 12 randomly selected fish from each treatment were placed into a breeding tank. Each trial consisted of placing spawning trays into the breeding tanks to induce spawning behavior, followed by egg collection the next day. Fish were rested 2 days (absence of spawning trays in the tank) between trials. Total cumulative egg production, total number of eggs per breeding trial, percent viability (percentage of total eggs that were viable), percent hatchability (percentage of total viable eggs that hatched), and percent swim-up success (as percentage of viable eggs) were determined. Following the breeding trials, fish from each treatment were euthanized, weights and lengths were recorded, and individual fish were sexed under a dissecting microscope. The gonads of each fish were removed and weighed to determine the gonadosomatic index (GSI), calculated as (weight of the gonads in mg)/(weight of the total body in mg) × 100. Ovaries and testes dissected from adult fish were used to determine gonadal HSP70 expression as described for 60 dph fish.

3. Results 3.1. Sixty days post-hatch survival, sex ratios and gonad histology

3.1.1. Controls Histological evaluation of gonads from whole mount sections of control zebrafish at 60 dph showed a sex ratio of 29.4% female and 53.0% male (Fig. 1) with respective normal ovarian and testicular ultrastructure and progression of gametogenesis. In the ovaries, a greater part of the gonad tissue was

2.10. Statistical analysis All data were tested for normality using the Kolmogorov– Smirnov test. If assumptions of normality and equal variance held true, then two-way analysis of variance (ANOVA) was performed to assess if any interaction existed between EE and NP on the examined endpoints. The different levels of EE were categorized into EE0, EE1, and EE10 while levels of NP were categorized into NP0, NP10, and NP100, creating a 3 × 3 two-way ANOVA matrix (Table 1). If an interaction was detected, further one-way ANOVA followed by Tukey’s post-hoc test was conducted to assess the significance of these effects. If assumptions of normality and equal variance failed, non-parametric Kruskal–Wallis one-way ANOVA on ranks

Fig. 1. Sex ratios of 60 days post-hatch zebrafish exposed to 10 ␮g/l nonylphenol (NP10; n = 19), 100 ␮g/l nonylphenol (NP100; n = 16), 1 ng/l 17␣ethinylestradiol (EE1; n = 19), 1 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE1 + NP10; n = 14), 1 ng/l 17␣-ethinylestradiol + 100 ␮g/l nonylphenol (EE1 + NP100; n = 14), or solvent control (CON, 0.01% acetone, v/v; n = 17) from 2 to 60 days post-hatch. No discernable gonadal tissues were observed in groups exposed to 10 ng/l 17␣-ethinylestradiol (EE10), 10 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE10 + NP10), or 10 ng/l 17␣ethinylestradiol + 100 ␮g/l nonylphenol (EE10 + NP100). Significantly different from control using chi-square test: ** P < 0.001; *** P < 0.0005. Und: undifferentiated gonad.

386

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

Fig. 2. Ovarian oogenesis staging of 60 days post-hatch female zebrafish. Percent distribution of different stages of ovarian follicle development in zebrafish exposed from 2 to 60 days post-hatch to 10 or 100 ␮g/l nonylphenol (NP10; n = 11 or NP100; n = 16, respectively), 1 ng/l 17␣-ethinylestradiol (EE1; n = 17), 1 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE1 + NP10; n = 14), 1 ng/l 17␣-ethinylestradiol + 100 ␮g/l nonylphenol (EE1 + NP100; n = 12) or solvent control (CON; n = 5). Four replicate views were evaluated blinded at 200× magnification in each fish and results were calculated using the mean value from each (n) fish. Numerical values (mean ± S.E.M) for the percent of total at each stage of ovarian follicle development observed are indicated adjacent to the corresponding pie area. Significantly different from solvent control using one-way ANOVA followed by Tukey’s test: * P < 0.05.

made up of previtellogenic oocytes (81.7 ± 2.3%; Fig. 2), surrounded by oogonia and early oocyte stages (18.3 ± 2.3%) in the caudal and cranial peripheries (Fig. 3b). In histologically determined males, the majority showed a full spectrum of sperm cell differentiation stages (Fig. 3a). In addition to the identified male and female, 17.6% of the control fish were classified as having undifferentiated gonads, containing primordial germ cells with no discernable cells characteristic of either sex (Fig. 3c). 3.1.2. NP-only exposure Fish exposed to NP10 had 57.9% female, 26.3% male, and 15.8% undifferentiated, which was significantly different from the control group (P < 0.001; Fig. 1). In females, the ovaries contained mainly previtellogenic oocytes (64.3 ± 7.0%) with the rest being oogonia and early oocyte stages (35.7 ± 7.0%; Fig. 2). Among the histologically determined males exposed to NP10, differences in the level of testicular development was observed, with 60% showing the full array of sperm cell differentiation and 40% exhibiting only the early stages of differentiation (data not shown). Fish exposed to NP100 were 100% female, with the ovaries consisting of 38.8 ± 6.9% previtellogenic oocytes

Fig. 3. Representative hematoxylin and eosin-stained gonad sections of 60 days post-hatch zebrafish. (a) Testis of control male at 800× magnification, (b) ovary of control female at 320× magnification and (c) undifferentiated gonad (inside white box) of a fish exposed to 1 ng/l 17␣-ethinylestradiol from 2 to 60 days post-hatch at 800× magnification. Sc: spermatocyte; Sg: spermatogonia; M: mature sperm; PV: previtellogenic follicle; Oo: oogonia.

and 61.2 ± 6.9% oogonia (Fig. 2). The sex ratio and ovarian gametogenesis of the NP100 group were significantly different from the control (P < 0.0005 and <0.05, respectively). No evidence of ovotestes was present in either of the two NP-only treatments.

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

387

Table 2 Length, weight, and condition factor determined in 60 days post-hatch zebrafish exposed to 10 or 100 ␮g/l nonylphenol (NP10 or NP100, respectively), 1 or 10 ng/l 17␣-ethinylestradiol (EE1 or EE10), 1 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE1 + NP10), 1 ng/l 17␣-ethinylestradiol + 100 ␮g/l nonylphenol (EE1 + NP100), 10 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE10 + NP10), 10 ng/l 17␣-ethinylestradiol + 100 ␮g/l nonylphenol (EE10 + NP100), or solvent control (0.01% acetone, v/v) from 2 to 60 days post-hatch Treatment

Length (mm)

Control (n = 42) NP10 (n = 42) NP100 (n = 42) EE1 (n = 42) EE10 (n = 42) EE1 + NP10 (n = 42) EE1 + NP100 (n = 42) EE10 + NP10 (n = 42) EE10 + NP100 (n = 42)

12.5 13.2 13.0 12.7 9.1 13.3 12.3 9.5 10.0

± ± ± ± ± ± ± ± ±

0.3 0.3 0.4 0.2 0.2*** 0.4 0.4 0.2*** 0.2***

Weight (mg) 13.5 15.6 16.5 13.4 6.3 17.3 13.5 7.8 7.8

± ± ± ± ± ± ± ± ±

0.9 0.9 1.5 0.7 0.3*** 1.3 1.1 0.6*** 0.4***

Condition factor 0.66 0.64 0.78 0.66 0.84 0.69 0.68 0.84 0.76

± ± ± ± ± ± ± ± ±

0.01 0.01 0.14 0.04 0.03*** 0.01 0.01 0.03*** 0.02**

Data are mean ± S.E.M. Significantly different from control using Dunn’s multiple comparisons: ** P < 0.01 and *** P < 0.001.

3.1.3. EE-only exposure The sex ratio of EE1-exposed fish was significantly different when compared with the control (P < 0.0005), with the majority being female (89.4%), while 5.3% male and 5.3% undifferentiated made up the remaining population (Fig. 1). The ovaries were comprised mainly of oogonia (60.5 ± 6.4%) with 39.5 ± 6.4% previtellogenic oocytes, which was significantly different than the control (P < 0.05; Fig. 2). The single fish identified histologically as male had testes displaying only the early stages of sperm cell differentiation (data not shown). There were no discernable gonadal tissues present in fish exposed to 10 ng/l of EE (Fig. 1). No evidence of ovotestes was present in either of the two EE-only treatments. 3.1.4. Binary mixture exposure Fish exposed to EE1 + NP10 resulted in 100% female (Fig. 1), whose ovaries contained a roughly equal portion of oogonia and previtellogenic oocytes (56.8 ± 6.3 and 43.2 ± 6.3% respectively; Fig. 2). In the EE1 + NP10 exposure group, sex ratio and ovarian gametogenesis were significantly different from the control (P < 0.0005 and <0.05, respectively). Exposure to EE1 + NP100 produced 85.8% female, 7.1% male and 7.1% undifferentiated; this ratio was significantly different when compared with the control (P < 0.0005). The ovaries in this group contained 57.4 ± 6.2% oogonia and 42.6 ± 6.2% previtellogenic oocytes, which was significantly different than the control (P < 0.05). The one individual identified as having testes showed the full range of sperm cell differentiation stages. There were no discernable gonadal tissues present in fish exposed to treatment groups containing 10 ng/l of EE (EE10, EE10 + NP10 and EE10 + NP100; Fig. 1). No evidence of ovotestes was present in any of the binary mixture treatments. 3.2. Sixty days post-hatch length, weight, and condition factor At 60 dph, only fish in treatments containing 10 ng/l of EE (EE10, EE10 + NP10 and EE10 + NP100), the same groups that showed no discernable gonadal tissues, exhibited decreased

length and weight, but higher condition factor, compared with control fish (Table 2; P < 0.001 for all except condition factor of EE10 + NP100, where P < 0.01). The remaining treatments were not significantly different in length, weight, or condition factor compared with the control (Table 2). 3.3. Sixty days post-hatch vitellogenin induction At 60 dph, the gender of fish used for VTG analyses was not determined due to the species’ lack of prominent sexual dimorphism at this developmental stage. Therefore, the VTG concentrations presented here are mean values of combined genetic male and female fish. Two-way ANOVA indicated that both NP and EE alone had a significant effect on 60 dph VTG induction (P < 0.0005 and <0.0001, respectively). Furthermore,

Fig. 4. Whole body vitellogenin concentration of 60 days post-hatch zebrafish exposed to 10 or 100 ␮g/l nonylphenol (NP10 or NP100, respectively; n = 10), 1 or 10 ng/l 17␣-ethinylestradiol (EE1 or EE10, respectively; n = 10), 1 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE1 + NP10; n = 9), 1 ng/l 17␣ethinylestradiol + 100 ␮g/l nonylphenol (EE1 + NP100; n = 10), 10 ng/l 17␣ethinylestradiol + 10 ␮g/l nonylphenol (EE10 + NP10; n = 11), 10 ng/l 17␣ethinylestradiol + 100 ␮g/l nonylphenol (EE10 + NP100; n = 10), or solvent control (CON, 0.01% acetone, v/v; n = 10) from 2 to 60 days post-hatch. Data are mean ± S.E.M. (a) Significant difference between NP100 and control (P < 0.05); (b) significant difference between EE1 + NP10 and EE1 (P < 0.001); (c) significant difference between EE10 + NP100 and EE10 + NP10 (P < 0.05); (d) significant difference between EE10 + NP100 and EE10 (P < 0.001).

388

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

there was a significant interaction between EE and NP on VTG induction (P < 0.0001). Subsequently, one-way ANOVAs were conducted on each subgroup to examine the effect of different levels of NP (NP0, NP10 or NP100) at each level of EE (EE0, EE1 or EE10) on VTG induction at 60 dph (Fig. 4). In the absence of EE (EE0), whole body VTG levels in the NP-100 exposed, but not NP10-exposed, zebrafish were significantly greater than the control (P < 0.05; Fig. 4). In the presence of 1 ng/l EE, the addition of 10 ␮g/l NP (EE1 + NP10) increased VTG significantly (P < 0.05) when compared to EE1 alone. However, with the addition of 100 ␮g/l NP, the VTG level in EE1 + NP100 was not significantly different from EE1 (Fig. 4). Analysis of the EE10-exposed groups showed that VTG induction in EE10 + NP100 was significantly lower than EE10 + NP10 (P < 0.05) as well as EE10 (P < 0.001) exposure groups (Fig. 4). One-way ANOVA comparing EE-only exposure groups with the control indicated that VTG induction in EE1 was not significantly different than the control, while EE10 had a significantly higher VTG induction than the control (P < 0.001). 3.4. Sixty days post-hatch whole body HSP70 induction Two-way ANOVA indicated that individual treatments of EE or NP had no significant effect on the 60 dph whole body HSP70 protein expression. As well, there was no interaction detected between EE and NP on HSP70 expression (Fig. 5). 3.5. Adult breeding experiments After a recovery period of 6 months in clean water, zebrafish were assessed for their reproductive fitness at 240 dph. A suc-

Fig. 5. Whole body heat shock protein 70 (HSP70) expression determined using Western blotting of 60 days post-hatch zebrafish exposed to 10 or 100 ␮g/l nonylphenol (NP10 or NP100, respectively), 1 or 10 ng/l 17␣-ethinylestradiol (EE1 or EE10, respectively), 1 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE1 + NP10), 1 ng/l 17␣-ethinylestradiol + 100 ␮g/l nonylphenol (EE1 + NP100), 10 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE10 + NP10), 10 ng/l 17␣-ethinylestradiol + 100 ␮g/l nonylphenol (EE10 + NP100), or solvent control (CON, 0.01% acetone, v/v) from 2 to 60 days post-hatch. (a) Representative immunoblot; (b) densitometry of immunoreactive bands.

cessful breeding event was defined as the production of one or more eggs. The highest percentage of successful breeding trials was in the control group with 95.6% (43/45; 43 successful breeding events out of 45 conducted, Fig. 6a), and the lowest percentage of successful trials was observed in the EE10 + NP10 group with 0% (0/45). The percentages of successful trials in other groups were as follows: NP10 (88.9%; 40/45), NP100 (88.9%; 40/45), EE1 (71.1%; 32/45), EE10 (40.0%; 18/45), EE1 + NP10 (86.7%; 39/45), EE1 + NP100 (86.7%; 39/45), and EE10 + NP100 (57.8%; 26/45). From the 45 breeding trials conducted in each treatment, it appeared that fish from groups exposed to 10 ng/l EE, alone or in combination, were the most adversely affected with respect to cumulative egg production. Zebrafish exposed to EE10, EE10 + NP10, and EE10 + NP100 had the three lowest cumulative egg productions, with 1965, 0, and 1808 eggs, respectively (Fig. 6b). Control, NP10, and EE1 + NP100 groups had the three highest cumulative egg productions, with 10,752, 11,275, and 12,754 eggs, respectively. Data from the remaining groups include EE1 with 4926 eggs, NP100 with 6705 eggs, and EE1 + NP10 with 8545 eggs (Fig. 6b). Two-way ANOVA of the mean number of eggs produced per breeding trial indicated that EE-only treatments significantly decreased egg production (P < 0.0001). Although treatments of NP alone did not significantly affect egg production, a significant interaction between NP and EE on egg production was detected (P < 0.0001). Subsequent one-way ANOVAs conducted on the EE0 subgroups indicated that neither NP10 or NP100 exposure had a significant impact on egg production (Fig. 6c). In 1 ng/l EE-exposed fish, the addition of 10 ␮g/l NP (EE1 + NP10) did not significantly influence egg production when compared with EE1, however the addition of 100 ␮g/l NP (EE1 + NP100) increased egg production significantly when compared with EE1 (P < 0.01; Fig. 6c). With the EE10-containing groups, egg production from EE10 + NP10 was significantly decreased when compared with EE10 and EE10 + NP100 (P < 0.01 for both), while no significant difference was detected between EE10 and EE10 + NP100. One-way ANOVA comparing the control with EE-only groups indicated that both EE1 and EE10 had significantly reduced mean egg production (P < 0.001 for both; Fig. 6c). One of the criteria for two-way ANOVA is that the sample size of each treatment group must be greater than 1. Since zebrafish exposed to EE10 + NP10 did not produce any eggs during the 45 breeding trials conducted, there were no subsequent data on egg viability and hatchability, as well as swim-up success of the F1 generation. Consequently, two-way ANOVA was not performed on these endpoints. However, using one-way ANOVA, it was demonstrated that fish previously exposed to 10 ␮g/l of NP did not exhibit any significant differences in viability (78.7 ± 2.0%), hatchability (90.6 ± 1.9%) or F1 swim-up success (80.6 ± 2.2%) when compared with the control (80.1 ± 2.4% viability, 84.2 ± 3.0% hatchability, and 89.9 ± 1.2% swim-up success; Fig. 6d). Fish exposed to EE1 had a significant decrease in egg viability (63.6 ± 4.2%, P < 0.01) but no differences in hatchability (85.5 ± 3.6%) or swim-up

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

389

Fig. 6. (a) Percent of successful breeding trials, (b) total cumulative number of eggs spawned, (c) mean number of eggs produced per breeding trial, and (d) percent viable eggs (viable eggs), percent hatch (hatchability), and percent F1 swim-up success (swim-up) determined using adult (240 dph) zebrafish previously exposed to 10 or 100 ␮g/l nonylphenol (NP10 or NP100, respectively), 1 or 10 ng/l 17␣-ethinylestradiol (EE1 or EE10), 1 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE1 + NP10), 1 ng/l 17␣-ethinylestradiol + 100 ␮g/l nonylphenol (EE1 + NP100), 10 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE10 + NP10), 10 ng/l 17␣ethinylestradiol + 100 ␮g/l nonylphenol (EE10 + NP100), or solvent control (CON, 0.01% acetone, v/v) from 2 to 60 days post-hatch. Two-way ANOVA indicated a significant interaction between EE and NP on mean number of eggs produced (P < 0.0001). Subsequent one-way ANOVA, followed by Tukey’s test, indicated: (a) significant difference between EE1 + NP100 and EE1 (P < 0.01); (b) significant difference between EE10 + NP10 and EE10 (P < 0.01); (c) significant difference between EE10 + NP10 and EE10 + NP100 (P < 0.01). Significantly different from solvent control using one-way ANOVA followed by Tukey’s test: ** P < 0.01; *** P < 0.001.

success (81.8 ± 3.6%) when compared with the control. Fish in the NP100 and the EE1 + NP10 exposure groups showed significant reductions in F1 swim-up success (74.7 ± 2.8 and 73.2 ± 2.5% respectively, P < 0.001) but no differences in viability (75.5 ± 2.8 and 71.6 ± 3.2% respectively) and hatchability (78.0 ± 2.7 and 88.6 ± 2.4% respectively) of eggs as compared to the control. Groups EE1 + NP100 and EE10 + NP100 produced fish that showed significant reductions in egg viability (63.2 ± 3.8 and 62.6 ± 3.7%, respectively, P < 0.001) and F1 swim-up success (71.6 ± 3.4 and 69.9 ± 4.5%, respectively, P < 0.001) but not hatchability (85.6 ± 2.8 and 84.9 ± 3.1%, respectively) compared with the control. Fish exposed to EE10 showed significant reductions in egg viability (51.4 ± 7.1%, P < 0.001), hatchability (64.4 ± 7.1%, P < 0.01) and F1 swimup success (59.1 ± 4.0%, P < 0.001) when compared with the control.

3.6. Adult sex ratio At the end of the reproductive trials, the gender of fish from all treatments were determined under a dissecting microscope. The control group had 58.7% males and 41.3% females (Fig. 7). The NP10 exposure group contained 53.8% males and 46.2% females, while NP100 contained 75.8% males and 24.2% females, which was statistically significant when compared with the control (P < 0.05). The EE1 exposure group consisted of 61.2% males and 38.8% females, whereas the EE10 exposure group had a predominantly male population with 92.2% males and 7.8% females (P < 0.0005). The EE1 + NP10 exposure group was comprised of 70.0% males and 30.0% females, and EE1 + NP100 contained 64.3% males and 35.7% females. Neither of these groups had significantly different sex ratios when compared with the control. Both of

390

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

the EE10-containing mixture exposure groups had a primarily male population with EE10 + NP10 having 98.0% males and 2.0% females (significantly different from the control, P < 0.0005), and EE10 + NP100 consisting of 93.1% males and 6.9% females (significantly different from the control, P < 0.0005). 3.7. Adult length, weight, condition factor and gonadosomatic index

Fig. 7. Sex ratios of adult zebrafish previously exposed to 10 or 100 ␮g/l nonylphenol (NP10; n = 52 or NP100; n = 33), 1 or 10 ng/l 17␣-ethinylestradiol (EE1; n = 64 or EE10; n = 63), 1 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE1 + NP10; n = 40), 1 ng/l 17␣-ethinylestradiol + 100 ␮g/l nonylphenol (EE1 + NP100; n = 42), 10 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE10 + NP10; n = 49), 10 ng/l 17␣-ethinylestradiol + 100 ␮g/l nonylphenol (EE10 + NP100; n = 72), or solvent control (CON, 0.01% acetone, v/v; n = 63) from 2 to 60 days post-hatch. Significantly different from control using chisquare test: * P < 0.05; *** P < 0.0005.

Fish in the highest binary exposure group (EE10 + NP100) had significantly lower weights (P < 0.05) and shorter lengths (P < 0.05), but no differences in condition factor when compared with the control (Table 3a). Fish in the NP100 exposure group were not significantly different in lengths or weights, but had a significantly higher condition factor when compared with the control (P < 0.05). The remaining treatments exhibited no significant differences in length, weight, or condition factor. With the exception of significantly lower GSI in EE10 adult males, no differences in GSI from any treatments were observed when compared with the control (Table 3b). 3.8. Adult gonadal HSP70 expression

Table 3 (a) Length, weight, and condition factor and (b) gonadosomatic index (GSI) determined in adult zebrafish previously exposed to 10 or 100 ␮g/l nonylphenol (NP10 or NP100, respectively), 1 or 10 ng/l 17␣-ethinylestradiol (EE1 or EE10), 1 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE1 + NP10), 1 ng/l 17␣-ethinylestradiol + 100 ␮g/l nonylphenol (EE1 + NP100), 10 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE10 + NP10), 10 ng/l 17␣ethinylestradiol + 100 ␮g/l nonylphenol (EE10 + NP100), or solvent control (0.01% acetone, v/v) from 2 to 60 days post-hatch Treatment

Length (mm)

(a) Control (n = 63) NP10 (n = 52) NP100 (n = 33) EE1 (n = 67) EE10 (n = 65) EE1 + NP10 (n = 39) EE1 + NP100 (n = 39) EE10 + NP10 (n = 49) EE10 + NP100 (n = 72)

39.6 39.2 39.3 39.1 38.1 40.1 39.5 39.9 37.6

± ± ± ± ± ± ± ± ±

0.5 0.6 0.4 0.4 0.6 0.4 0.3 0.3 0.4*

Weight (mg) 616 596 631 580 535 649 634 580 484

± ± ± ± ± ± ± ± ±

34 39 29 31 43 33 28 22 24*

There were no significant differences in the levels of gonadal HSP70 expression among treatments in adult zebrafish (Fig. 8).

Condition factor 0.94 0.93 1.02 0.93 0.87 0.98 1.01 0.91 0.87

Treatment

Female (n)

Male (n)

(b) Control NP10 NP100 EE1 EE10 EE1 + NP10 EE1 + NP100 EE10 + NP10 EE10 + NP100

0.179 ± 0.011 (26) 0.150 ± 0.016(24) 0.139 ± 0.019 (8) 0.162 ± 0.012 (26) 0.167 ± 0.053 (5) 0.189 ± 0.018 (11) 0.179 ± 0.016 (12) N/A 0.151 ± 0.028 (5)

0.013 0.011 0.015 0.011 0.010 0.014 0.013 0.014 0.011

± ± ± ± ± ± ± ± ±

± ± ± ± ± ± ± ± ±

0.01 0.02 0.02* 0.02 0.02 0.02 0.02 0.02 0.02

0.001 (37) 0.001 (28) 0.001 (25) 0.001 (38) 0.001 (58)* 0.001 (28) 0.001 (27) 0.001 (47) 0.001 (67)

Data are mean ± S.E.M. Significantly different from control: * P < 0.05.

Fig. 8. Gonadal heat shock protein 70 (Hsp70) expression of adult zebrafish. (a) Representative Western blot of Hsp70 expression in adult zebrafish gonadal tissues, (b) gonadal heat shock protein 70 expression in adult zebrafish previously exposed to 10 or 100 ␮g/l nonylphenol (NP10 or NP100, respectively), 1 or 10 ng/l 17␣-ethinylestradiol (EE1 or EE10), 1 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE1 + NP10), 1 ng/l 17␣-ethinylestradiol + 100 ␮g/l nonylphenol (EE1 + NP100), 10 ng/l 17␣-ethinylestradiol + 10 ␮g/l nonylphenol (EE10 + NP10), 10 ng/l 17␣ethinylestradiol + 100 ␮g/l nonylphenol (EE10 + NP100), or solvent control (CON, 0.01% acetone, v/v) from 2 to 60 days post-hatch. Data are mean optical density (OD) ± S.E.M.

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

4. Discussion It is generally assumed in aquatic ecotoxicology investigations that mixtures of xenoestrogens act in an additive manner. The major finding of the present study was that NP, a weak ER agonist, can behave both additively and non-additively when in combination with a potent ER agonist, EE, at environmentally relevant concentrations. This was particularly evident with VTG induction, a direct gene expression product following ER activation, in 60 dph zebrafish. Utilizing two-way ANOVA, it was observed that an interaction existed between EE and NP on VTG induction at 60 dph. Upon closer examination with oneway ANOVA on the effect of different levels of NP (NP0, NP10, and NP100) at each level of EE (EE0, EE1, and EE10), it was found that at 1 ng/l EE, the addition of 10 ␮g/l NP (EE1 + NP10) increased the VTG induction significantly when compared to EE1 alone, exhibiting additivity of effects as neither EE1 nor NP10 alone resulted in significant VTG induction when compared with the control. However, in the EE1 + NP100 exposure group, where the NP concentration increased from 10 to 100 ␮g/l while keeping the EE concentration constant at 1 ng/l, the level of VTG observed was significantly lower than that of EE1 + NP10, and statistically insignificant when compared with EE1. This demonstrated that during exposure of zebrafish to 1 ng/l EE, lower levels of NP (NP10) appeared to act additively with EE in terms of VTG induction, while higher levels of NP (NP100) appeared to antagonize the action of EE, inducing lower levels of VTG. The antagonistic effect of NP was even more pronounced with mixture groups containing the higher EE concentration (10 ng/l). Fish exposed to 10 ng/l of EE + 100 ␮g/l of NP had a significantly lower VTG level when compared with the EE10 + NP10 group and the EE10 alone group. One possible explanation for the observed non-additivity at certain combinations of EE + NP is that the EE and NP molecules are simply adhering to the mass action law in receptor binding. Using environmentally relevant concentrations where NP generally occurs at approximately ≥1000 times greater concentrations than EE, the higher concentration of NP could theoretically allow NP a better chance of competing with EE for the available ER binding sites. Consequently, in binary mixtures of EE + NP at environmentally relevant levels, the addition of NP would allow them to bind to some of the ER that would otherwise have been occupied by EE if EE was acting alone. The increase in NP concentration would therefore increase the proportion of ERs bound to NP. Since NP has been shown to be a weak or partial ER agonist, meaning that it binds to ER but produces a diminished response (e.g. VTG induction) compared to a full agonist like EE, the lower overall response induced by the binary mixture of NP and EE would be due to NP’s weaker estrogenicity in comparison to EE. Therefore, NP may not be acting as an ER antagonist in the classical sense, i.e. binding to the ER and inhibiting ER-dependent responses. However, by its very nature as a weak or partial ER agonist, it may be lowering ERdependent responses in binary mixtures of NP and EE where its concentration is high enough that it can displace EE binding to the ER, resulting in, in this case, lower VTG induction than when EE was acting alone. This hypothesis also allows us

391

to explain the additive effect at a lower EE + NP concentration (EE1 + NP10), since at these lower EE + NP concentrations, ER are not as saturated, allowing all EE and NP molecules to bind and elicit each compound’s respective level of estrogenicity. It is important to note however that the decreased VTG associated with increasing NP concentration in treatments containing 10 ng/l of EE (EE10, EE10 + NP10 and EE10 + NP100) is relative within those groups: the addition of NP to 10 ng/l of EE did not decrease VTG levels lower than that observed in groups containing 1 ng/l of EE (EE1, EE1 + NP10 and EE1 + NP100). There are few studies that have examined in vivo xenoestrogen mixture toxicity in fish. Nonylphenol was reported to act additively in binary mixtures with E2 , at concentrations below their individual LOECs, on VTG induction in juvenile female rainbow trout (Oncorhynchus mykiss) (Thorpe et al., 2001). Despite possible differences in species sensitivity to NP, lifestage sensitivity (3 months old versus newly hatched), sex (female versus mixed population) and duration of exposure (14 versus 60 days), the result by Thorpe et al. (2001) is in agreement with our low EE + NP (EE1 + NP10) data where exposure to each compound alone did not induce significant increases in VTG, while combination of the two resulted in a significant increase. Recent studies that have examined xenoestrogen mixture toxicity in other species or tissues have also reported departure from additivity of responses. Xie et al. (2005) utilized a rainbow trout VTG assay to evaluate the estrogenicity of four herbicides, two alkylphenol ethoxylate-containing surfactants, and binary mixtures of herbicides with the surfactants. They observed that 2,4-D alone displayed estrogenic activity via VTG induction and in binary mixtures with target prospreader activator (TPA), an alkylphenol ethoxylate-containing surfactant, exhibited greater than additive VTG response at the lowest concentrations tested, but a less than additive response at the highest combined concentrations. However, the results are complicated by the fact that 2,4-d contains several dioxin-like compounds which may bind to aryl hydrocarbon receptors (AhR) and interfere with the overall estrogenic response. In addition, the surfactant TPA may contain components other than alkyphenol ethoxylate that could affect the observed estrogenic response. Another recent study (Rajapakse et al., 2004) utilized the in vitro E-SCREEN assay to evaluate mixture toxicity of six xenoestrogens: E2 , EE, NP, octylphenol (OP), genistein and bisphenol A. It was demonstrated that the presence of NP and OP was associated with the antagonism observed in five- and six-component mixtures. It should be noted, however, that the endpoint measured via ESCREEN may not necessarily be exclusively estrogenic as the number of cancer cells produced within a given amount of time takes into account the number of cells produced through ER activation as well as cells lost through cytotoxicity. The authors reported that higher concentrations of NP and OP were accompanied by a marked decrease in cell numbers. The decrease may be attributed to cytotoxicity, and may or may not be ER-mediated as there are numerous ways in which chemicals can exert growthrestricting or cytotoxic effects (Rajapakse et al., 2004). An additional factor to consider when evaluating the nature of interaction(s) between compounds in mixture toxicity studies

392

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

is that the type of interaction may be dependent on the particular biological response(s) measured. In particular, responses closely linked with gene expression, such as VTG induction, will likely produce different patterns of mixture toxicity when compared to more complex, integrative responses such as reproduction. A similar trend of non-additivity in the present study was observed in certain responses determined in the adult breeding experiments. Specifically, exposure to NP100 in the presence of EE1 abolished the decreased mean egg production observed with EE1 alone. Similarly, in the EE10 exposure groups, the addition of 100 ␮g/l NP (EE10 + NP100) resulted in egg production that was not significantly different from EE10 alone. These results collectively suggest that lower levels of NP (NP10) may have acted additively in the presence of 10 ng/l EE, while higher levels of NP (NP100) did not further reduce the mean egg production, instead, resulted in mean egg production that was similar to EE10. Results from the breeding trials also highlighted the toxicity of early life stage exposure to 10 ng/l of EE in zebrafish. After a 6 month period of depuration, there was a persistent impact on several reproductive parameters. Treatments containing 10 ng/l of EE (EE10, EE10 + NP10 and EE10 + NP100) resulted in the lowest cumulative number of eggs produced during the duration of the breeding trial, and the same three groups exhibited the lowest percentages of successful breeding trials. Other studies investigating breeding success in zebrafish exposed during sexual differentiation to environmentally relevant concentrations of xenoestrogens have reported similar reductions in reproductive fitness. Hill and Janz (2003) exposed zebrafish from 2 to 60 dph to nominal concentrations of 10–100 ␮g/l of NP or 1–100 ng/l of EE, then raised in clean water from 60 to 120 dph, and reported that the NP–100 ␮g/l group had decreased egg hatchability and swim-up success, while EE–10 ng/l resulted in decreased egg viability and hatchability as well as swim-up success. Another study reported that the exposure of zebrafish embryos to nominal concentrations of 10 or 25 ng/l of EE until 90 days post-fertilization (dpf) and allowed to recover in clean water for 5 months, resulted in a reduced number of spawning females as well as reduced egg production (Van den Belt et al., 2003). Fenske et al. (2005) evaluated the reproductive capacities of zebrafish exposed to 3 ng/l EE from either 0–42 dpf followed by 76 days depuration or 0–118 dpf followed by 58 days depuration. They reported that there were no significant effects on the reproductive fitness of the fish exposed from 0 to 42 dpf, which differs from present study where exposure to a lower EE concentration (1 ng/l) from 2 to 60 dph, followed by 6 months of depuration, resulted in a significant decrease in egg viability. However, this may be explained by the difference in the duration of exposure, and is supported by their finding that zebrafish exposed for a longer period, from 0 to 118 dpf, had decreased egg production and fertilization success (Fenske et al., 2005). This further indicates the need to account for the length of exposure in assessing the effects of xenoestrogens, particularly when overlapping the period of sexual differentiation and gametogenesis. Other fish models such as medaka (Oryzias latipes) and fathead minnow (Pimephales promelas) have also been utilized in

similar investigations. Medaka exposed to 0.2 or 2 ng/l of EE from 2 to 5 dph until sexual maturity (between 4 and 6 months of age) displayed normal mating behavior and reproductive success, while males exposed to 10 ng/l of EE exhibited suppressed reproductive behavior, and females had poor reproductive success (Balch et al., 2004). In fathead minnow exposed to 0.2 or 1.0 ng/l of EE from 0 to 301 dph, no effects on female egg production were reported (Lange et al., 2001). The variation seen in estrogenic sensitivity highlights the importance of taking into account species differences when evaluating the effects of xenoestrogen exposure in fish. The results from the reproduction experiments in the present study were partly explained upon subsequent determination of the adult sex ratio. The EE10 + NP10 group had only one female at the end of the reproductive trials, which most likely contributed to the observed lack of egg production from that group. Two other groups that experienced significantly decreased mean egg production per breeding trial when compared with the control also contained lower numbers of females, 7.8% (n = 5) in EE10 and 6.9% (n = 5) in EE10 + NP100, when examined after completion of the reproductive studies. Interestingly, these three groups with the lowest female:male ratios and lowest mean egg production per breeding trial were the same groups that did not present any visible gonadal tissues during the 60 dph histological examination. The EE1-exposed fish also produced a significantly lower mean number of eggs per breeding trial, but the sex ratio was not significantly different from that of the control. The sex ratios of all treatment groups, excluding the EE10containing groups, which had no discernable gonadal tissues, were all skewed towards more females at the 60 dph histological examination. However, histological examination after the adult breeding trials revealed that the sex ratios of these groups did not deviate significantly from the control. From these observations, it appears that when zebrafish are exposed during early development to concentrations of xenoestrogens that are below 10 ng/l EE, namely the groups EE1, NP10, EE1 + NP10, and EE1 + NP100, the effects of xenoestrogen exposure on sexual differentiation are reversible following a period of depuration. Other studies have reported a similar trend, where sex ratios of zebrafish exposed to xenoestrogens during early development were skewed towards more female, but more males developed after cessation of xenoestrogen exposures (Hill and Janz, 2003; Fenske et al., 2005). The plasticity of zebrafish gonadal development was demonstrated through the reversal of sex ratio seen with the adult zebrafish after being exposed from 2 to 60 dph to various concentrations of EE, NP and EE + NP and allowed a 6 month depuration. In addition, upon measuring the adult GSI, with the exception of the significantly lowered GSI in EE10 adult males, no differences in GSI from any treatments were observed when compared with the control. Despite this apparent recovery, further evidence of reduced reproductive fitness, other than the aforementioned decreased daily and overall egg production, were observed. Eggs produced by fish in groups EE10, EE1 + NP100 and EE10 + NP100 experienced decreased viability when compared with the control. Decreased hatchability in EE10 was also observed. As well, the swim-up success of the F1

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

generation from groups EE10, EE1 + NP10, EE1 + NP100 and EE10 + NP100 was significantly reduced. This illustrates that exposure to xenoestrogens during the critical period of gonadal development and sexual differentiation, even after an extended period of depuration, may have irreversible effects on the reproductive system. Further research is warranted to elucidate the mechanism(s) of both the cause of reproductive impairment and transgenerational effects. Another biochemical endpoint examined in the current study was the expression of HSP70 in whole body juveniles and adult gonads to investigate the feasibility of correlating HSP70 responses with estrogenic exposures in zebrafish. There were no significant differences between treatments in the level of whole body HSP70 expression in the 60 dph fish. Although some studies have reported elevated levels of piscine HSP70 as well as other heat shock proteins associated with exposure to various environmental stressors including heavy metals, industrial effluents, pesticides, and polycyclic aromatic hydrocarbons, it is noteworthy that the majority of studies evaluated the response in cell lines, primary cell cultures or specific tissues and not of the whole animal (Basu et al., 2002). Other studies have indicated that the HSP response can differ among tissues (Smith et al., 1999; Rabergh et al., 2000; Yoo and Janz, 2003), which may help explain the lack of significant differences in the whole body HSP70 levels between the groups in this study since the measurement accounted for not only specific tissue types but the whole animal. There were also no significant differences between treatments in the gonadal HSP70 expression of the adult zebrafish at the end of the reproductive study. It is likely that 6 months of depuration after the 60 day exposure period was sufficient time for HSP70 to return to basal levels. Based on our results, we would argue that whole body HSP70 does not appear to be a viable biomarker for chronic xenoestrogen exposure due to possible variations among tissues. The length of exposure for the present study was set at 60 days. Previous studies have demonstrated that 60 dph is sufficient time for the complete sexual differentiation in zebrafish. Takahashi (1977) reported that complete sex reversal and testicular formation occurs by 40 dph. Uchida et al. (2002) also reported a gonad transformation period 3–4 weeks after hatching, with the total disappearance of all oocytes in genetic males occurring by the end of 4 weeks. However, in the current study, 18% of the control fish at 60 dph possessed undifferentiated gonads. Andersen et al. (2004) reported similar results when studying the effects of the anti-estrogen ZM 189,156 and the aromatase inhibitor fadrozole on juvenile zebrafish. They described that at 60 dph, the water and solvent control groups had 14 and 24% of fish possessing undifferentiated gonads, respectively. Other studies have also reported the presence of undifferentiated gonads in control zebrafish at 60 dph (Hill and Janz, 2003; Orn et al., 2003). The variation observed in timing of the gonad maturation period between the studies may be attributed to strain differences and/or the different rearing conditions of each respective study (Orn et al., 2003). Maack et al. (2003) suggested that stocking density, feeding conditions, social factors and water temperature, and not strain differences, may affect the timing of sexual development in zebrafish. It is also possible that the par-

393

ticular batch of fish used in the present study might not have been fully developed at 60 dph, as Maack et al. (2003) observed that high inter-individual variability exists in the timing of sex differentiation within each strain with no obvious correlation to body mass. This appears to be the most likely explanation for the present study as the body mass of the control fish at 60 dph was similar to the controls in Van den Belt et al. (2003), while less than the controls in Andersen et al. (2004), Hill and Janz (2003), and Orn et al. (2003), with no apparent correlation to each respective level of undifferentiated gonads. In the present study, we also observed the lack of visible gonad tissues at 60 dph in fish from exposure groups containing 10 ng/l of EE (EE10, EE10 + NP10 and EE10 + NP100). Andersen et al. (2003) also observed 60 dph zebrafish with no visible gonads when exposed to 15.4 ± 1.4 ng/l EE from hatch to 60 dph. As well, Van den Belt et al. (2003) reported a concentration-dependent increase in the number of fish with no discernable gonads when exposed to 0.1, 1, 10 or 25 ng/l of EE from hatch until 3 months post-hatch. One possible explanation, as mentioned above, could be that the energy required for VTG production is quite high, thus the energy normally devoted to gonad development may be diverted to VTG synthesis. Consequently, groups exhibiting high VTG induction would yield more individuals with small or underdeveloped gonads. These results suggest that the delay in the maturation process associated with EE exposures in zebrafish is both concentration- and duration-dependent. In summary, the results of the current study suggest that mixtures of weak and potent ER agonists, a scenario most likely experienced by wild fish populations, may produce certain effects that deviate from the assumption of simple additivity. However, such non-additivity may be complicated by the relative concentrations of each component of the mixture, and by the specific biological response being determined. Acknowledgements This study was funded by the Natural Sciences and Engineering Research Council of Canada (NSERC). References Aherne, G.W., Briggs, R., 1989. The relevance of the presence of certain synthetic steroids in the aquatic environment. J. Pharm. Pharmacol. 41, 735–736. Allen, Y., Matthiessen, P., Scott, A.P., Haworth, S., Feist, S., Thain, J.E., 1999. The extent of oestrogenic contamination in the UK estuarine and marine environments—further surveys of flounder. Sci. Total Environ. 233, 5–20. Andersen, L., Holbech, H., Gessbo, A., Norrgren, L., Petersen, G.I., 2003. Effects of exposure to 17alpha-ethinylestradiol during early development on sexual differentiation and induction of vitellogenin in zebrafish (Danio rerio). Comp. Biochem. Physiol. 34C, 365–374. Andersen, L., Kinnberg, K., Holbech, H., Korsgaard, B., Bjerregaard, P., 2004. Evaluation of a 40 day assay for testing endocrine disrupters: effects of an anti-estrogen and an aromatase inhibitor on sex ratio and vitellogenin concentrations in juvenile zebrafish (Danio rerio). Fish Physiol. Biochem. 30, 257–266. Arcand-Hoy, L.D., Nimrod, A.C., Benson, W.H., 1998. Endocrine modulating substances in the environment—estrogenic effects of pharmaceutical products. Int. J. Toxicol. 17, 139–158. Ashfield, L.A., Pottinger, T.G., Sumpter, J.P., 1998. Exposure of female juvenile rainbow trout to alkylphenolic compounds results in modifica-

394

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395

tions to growth and ovosomatic index. Environ. Toxicol. Chem. 17, 679– 686. Balch, G.C., Mackenzie, C.A., Metcalfe, C.D., 2004. Alterations to gonadal development and reproductive success in Japanese medaka (Oryzias latipes) exposed to 17alpha-ethinylestradiol. Environ. Toxicol. Chem. 23, 782–791. Basu, N., Todgham, A.E., Ackerman, P.A., Bibeau, M.R., Nakano, K., Schulte, P.M., Iwama, G.K., 2002. Heat shock protein genes and their functional significance in fish. Gene 295, 173–183. Belfroid, A.C., Van der Horst, A., Vethaak, A.D., Schafer, A.J., Rijs, G.B., Wegener, J., Cofino, W.P., 1999. Analysis and occurrence of estrogenic hormones and their glucuronides in surface water and waste water in The Netherlands. Sci. Total Environ. 225, 101–108. Bennie, D.T., 1999. Review of the environmental occurrence of alkylphenols and alkylphenol ethoxylates. Water Qual. Res. J. Can. 34, 79–122. Brian, J.V., Harris, C.A., Scholze, M., Backhaus, T., Booy, P., Lamoree, M., Pojana, G., Jonkers, N., Runnalls, T., Bonfa, A., Marcomini, A., Sumpter, J.P., 2005. Accurate prediction of the response of freshwater fish to a mixture of estrogenic chemicals. Environ. Health Perspect. 113, 721–728. Colborn, T., vom Saal, F.S., Soto, A.M., 1993. Developmental effects of endocrine-disrupting chemicals in wildlife and humans. Environ. Health Perspect. 101, 378–384. Danzo, B.J., 1997. Environmental xenobiotics may disrupt normal endocrine function by interfering with the binding of physiological ligands to steroid receptors and binding proteins. Environ. Health Perspect. 105, 294–301. Desbrow, C., Routledge, E.J., Brighty, G.C., Sumpter, J.P., Waldock, M., 1998. Identification of estrogenic chemicals in STW effluent. 1. Chemical fractionation and in vitro biological screening. Environ. Sci. Technol. 32, 1549–1558. Fenske, M., Maack, G., Schafers, C., Segner, H., 2005. An environmentally relevant concentration of estrogen induces arrest of male gonad development in zebrafish, Danio rerio. Environ. Toxicol. Chem. 24, 1088–1098. Folmar, L.C., Denslow, N.D., Rao, V., Chow, M., Crain, D.A., Enblom, J., Marcino, J., Guillette Jr., L.J., 1996. Vitellogenin induction and reduced serum testosterone concentrations in feral male carp (Cyprinus carpio) captured near a major metropolitan sewage treatment plant. Environ. Health Perspect. 104, 1096–1101. Gray, M.A., Metcalfe, C.D., 1997. Induction of testis-ova in Japanese medaka (Oryzias latipes) exposed to p-nonylphenol. Environ. Toxicol. Chem. 16, 1082–1086. Guengerich, F.P., 1990. Metabolism of 17 alpha-ethynylestradiol in humans. Life Sci. 47, 1981–1988. Guillette Jr., L.J., Gunderson, M.P., 2001. Alterations in development of reproductive and endocrine systems of wildlife populations exposed to endocrinedisrupting contaminants. Reproduction 122, 857–864. Guillette Jr., L.J., Gross, T.S., Masson, G.R., Matter, J.M., Percival, H.F., Woodward, A.R., 1994. Developmental abnormalities of the gonad and abnormal sex hormone concentrations in juvenile alligators from contaminated and control lakes in Florida. Environ. Health Perspect. 102, 680–688. Hill Jr., R.L., Janz, D.M., 2003. Developmental estrogenic exposure in zebrafish (Danio rerio). I. Effects on sex ratio and breeding success. Aquat. Toxicol. 63, 417–429. Iwama, G.K., Thomas, P.T., Forsyth, R.B., Vijayan, M.M., 1998. Heat shock protein expression in fish. Rev. Fish Biol. Fish. 8, 35–56. Janz, D.M., McMaster, M.E., Munkittrick, K.R., Van der Kraak, G., 1997. Elevated ovarian follicular apoptosis and heat shock protein-70 expression in white sucker exposed to bleached kraft pulp mill effluent. Toxicol. Appl. Pharmacol. 147, 391–398. Janz, D.M., McMaster, M.E., Weber, L.P., Munkittrick, K.R., Van der Kraak, G., 2001. Recovery of ovary size, follicle cell apoptosis, and HSP70 expression in fish exposed to bleached pulp mill effluent. Can. J. Fish. Aquat. Sci. 58, 620–625. Jobling, S., Sumpter, J.P., 1993. Detergent components in sewage effluent are weakly oestrogenic to fish: An in vitro study using rainbow trout (Oncorhynchus mykiss) hepatocytes. Aquat. Toxicol. 27, 361–372. Jobling, S., Sheahan, D., Osborne, J.A., Matthiessen, P., Sumpter, J.P., 1996. Inhibition of testicular growth in rainbow trout (Oncorhynchus mykiss) exposed to estrogenic alkyl-phenolic chemicals. Environ. Toxicol. Chem. 15, 194–202.

Khan, I.A., Thomas, P., 1998. Estradiol-17 beta and o,p’-DDT stimulate gonadotropin release in Atlantic croaker. Mar. Environ. Res. 46, 149–152. Lange, R., Hutchinson, T.H., Croudace, C.P., Siegmund, F., Schweinfurth, H., Hampe, P., Panter, G.H., Sumpter, J.P., 2001. Effects of the synthetic estrogen 17 alpha-ethinylestradiol on the life-cycle of the fathead minnow (Pimephales promelas). Environ. Toxicol. Chem. 20, 1216–1227. Maack, G., Segner, H., Tyler, C.R., 2003. Ontogeny of sexual differentiation in different strains of zebrafish (Danio rerio). Fish Physiol. Biochem. 28, 125–128. Naylor, G.C., Mierure, J.P., Weeks, J.A., Castaldi, F.J., Romano, R.R., 1992. Alkylphenol ethoxylates in the environment. J. Am. Oil Chem. Soc. 69, 695–703. Nimrod, A.C., Benson, W.H., 1996. Environmental estrogenic effects of alkylphenol ethoxylates. Crit. Rev. Toxicol. 26, 335–364. Orn, S., Holbech, H., Madsen, T.H., Norrgren, L., Petersen, G.I., 2003. Gonad development and vitellogenin production in zebrafish (Danio rerio) exposed to ethinylestradiol and methyltestosterone. Aquat. Toxicol. 65, 397–411. Rabergh, C.M., Airaksinen, S., Soitamo, A., Bjorklund, H.V., Johansson, T., Nikinmaa, M., Sistonen, L., 2000. Tissue-specific expression of zebrafish (Danio rerio) heat shock factor 1 mRNAs in response to heat stress. J. Exp. Biol. 203, 1817–1824. Rajapakse, N., Silva, E., Scholze, M., Kortenkamp, A., 2004. Deviation from additivity with estrogenic mixtures containing 4-nonylphenol and 4-tertoctylphenol detected in the E-SCREEN assay. Environ. Sci. Technol. 38, 6343–6352. Routledge, E.J., Sumpter, J.P., 1996. Estrogenic activity of surfactants and some of their degradation products assessed using a recombinant yeast screen. Environ. Toxicol. Chem. 15, 241–248. Ryan, J.A., Hightower, L.E., 1994. Evaluation of heavy-metal ion toxicity in fish cells using a combined stress protein and cytotoxicity assay. Environ. Toxicol. Chem. 13, 1231–1240. Smith, T.R., Tremblay, G.C., Bradley, T.M., 1999. Characterization of the heat shock protein response of Atlantic salmon (Salmo salar). Fish Physiol. Biochem. 20, 279–292. Soto, A.M., Sonnenschein, C., Chung, K.L., Fernandez, M.F., Olea, N., Serrano, F.O., 1995. The E-SCREEN assay as a tool to identify estrogens: an update on estrogenic environmental pollutants. Environ. Health Perspect. 103 (Suppl 7), 113–122. Sumpter, J.P., 1998. Xenoendorine disrupters-environmental impacts. Toxicol. Lett. 102–103, 337–342. Sumpter, J.P., Jobling, S., 1995. Vitellogenesis as a biomarker for estrogenic contamination of the aquatic environment. Environ. Health Perspect. 103 (Suppl 7), 173–178. Takahashi, H., 1977. Juvenile hermaphroditism in the zebrafish, Brachydanio rerio. Bull. Faculty Fish. Hokkaido Univ. 28, 57–65. Ternes, T.A., Stumpf, M., Mueller, J., Haberer, K., Wilken, R.D., Servos, M., 1999. Behavior and occurrence of estrogens in municipal sewage treatment plants.I. Investigations in Germany, Canada and Brazil. Sci. Total Environ. 225, 81–90. Thorpe, K.L., Hutchinson, T.H., Hetheridge, M.J., Scholze, M., Sumpter, J.P., Tyler, C.R., 2001. Assessing the biological potency of binary mixtures of environmental estrogens using vitellogenin induction in juvenile rainbow trout (Oncorhynchus mykiss). Environ. Sci. Technol. 35, 2476–2481. Thorpe, K.L., Cummings, R.I., Hutchinson, T.H., Scholze, M., Brighty, G., Sumpter, J.P., Tyler, C.R., 2003. Relative potencies and combination effects of steroidal estrogens in fish. Environ. Sci. Technol. 37, 1142–1149. Tyler, C.R., Jobling, S., Sumpter, J.P., 1998. Endocrine disruption in wildlife: a critical review of the evidence. Crit. Rev. Toxicol. 28, 319–361. Uchida, D., Yamashita, M., Kitano, T., Iguchi, T., 2002. Oocyte apoptosis during the transition from ovary-like tissue to testes during sex differentiation of juvenile zebrafish. J. Exp. Biol. 205, 711–718. Van den Belt, K., Verheyen, R., Witters, H., 2003. Effects of 17alphaethynylestradiol in a partial life-cycle test with zebrafish (Danio rerio): effects on growth, gonads and female reproductive success. Sci. Total Environ. 309, 127–137. Vijayan, M.M., Pereira, C., Kruzynski, G., Iwama, G.K., 1998. Sublethal concentrations of contaminant induce the expression of hepatic heat shock protein 70 in two salmonids. Aquat. Toxicol. 40, 101–108.

L.L. Lin, D.M. Janz / Aquatic Toxicology 80 (2006) 382–395 Weber, L.P., Janz, D.M., 2001. Effect of beta-naphthoflavone and dimethylbenz[a]anthracene on apoptosis and HSP70 expression in juvenile channel catfish (Ictalurus punctatus) ovary. Aquat. Toxicol. 54, 39– 50. Weber, L.P., Hill Jr., R.L., Janz, D.M., 2003. Developmental estrogenic exposure in zebrafish (Danio rerio). II. Histological evaluation of gametogenesis and organ toxicity. Aquat. Toxicol. 63, 431–446. Westerfield, M., 1995. The Zebrafish Book: A Guide for the Laboratory Use of Zebrafish Danio rerio, 3rd ed. University of Oregon Press, Eugene, OR, USA. Williams, J.H., Farag, A.M., Stansbury, M.A., Young, P.A., Bergman, H.L., Petersen, N.S., 1996. Accumulation of hsp70 in juvenile and adult rainbow

395

trout gill exposed to metal-contaminated water and/or diet. Environ. Toxicol. Chem. 15, 1324–1328. Williams, R.J., Johnson, A.C., Smith, J.J., Kanda, R., 2003. Steroid estrogens profiles along river stretches arising from sewage treatment works discharges. Environ. Sci. Technol. 37, 1744–1750. Xie, L., Thrippleton, K., Irwin, M.A., Siemering, G.S., Mekebri, A., Crane, D., Berry, K., Schlenk, D., 2005. Evaluation of estrogenic activities of aquatic herbicides and surfactants using an rainbow trout vitellogenin assay. Toxicol. Sci. 87, 391–398. Yoo, J.L., Janz, D.M., 2003. Tissue-specific HSP70 levels and reproductive physiological responses in fishes inhabiting a metal-contaminated creek. Arch. Environ. Contam. Toxicol. 45, 110–120.